摘要
我们针对肺癌的各种检查方法设计了以医院人群为基础的病例对照研究。应用Ｍａｎｔｅｌ－Ｈａｅｎｓｚｅｌ及非条件ｌｏｇｉｓｔｉｃ回归方程进行了单因素及多因素分析。结果表明：在研究对象中，吸烟史、家族史、肿瘤史、Ｘ线检查、ＣＴ检查与肺癌的关系在统计学上有显著意义。ＰＡＴ痰检、ＣＴ检查、Ｘ线检查、主要症状是影响肺癌临床诊断的主要因素。建立肺癌模型，回代结果显示：灵敏度为９３．９２％，特异度为７３．２８％，准确度９０．１９％。这对提高肺癌临床诊断水平，减少漏诊率有着积极的作用。
We conducted a case control study on the basis of inpatients for the examination methods of lung cancer.We applied Mantel Haenszel and nonconditional stepwise logistic model to simple factor and multivariate analysis. 526 cases of lung cancer were collected, at the same teme,116 controls were collected from the same hospital with nomalignant. The results showed that age, sex,and occupation are all significant relationship with the types of pathology for lung cancer (P<0 05);And smoking history,family history,tumor history,the findings X- rays and the findings of CT statictically significantly associate with lung cancer.The PAT sputum cytology,CT examination,X- rays examination and main symptom are all principal factors of effecting on the early diagnosis of lung cancer. We should alert on the people with the diagnostic positive of lung cancer.Therefore,the diagnostic computer system of lung cancer founded helps us to find lung cancer earlier and take the active treatment measures.
出处
《中国卫生统计》
CSCD
北大核心
1997年第5期11-14,共4页
Chinese Journal of Health Statistics
基金
辽宁省科学技术基金
关键词
肺肿瘤
诊断模型
Lung cancer The diagnostic computer system
